Question

    Many people will be familiar with automated writing through Gmail. Smart Reply provides brief answers to routine emails. If someone asks “Do you want to meet at 3 pm?”, Gmail offers one-click responses such as “Sure!”

The New Yorker’s John Seabrook recently described a more powerful version of this technology, called GPT-2, which can ably mimic (模拟) his magazine’s style. Such systems use a digital network of billions of artificial “neurons” (神经元) with virtual “synapses” — the connections between neurons — that strengthen as the network “learns”, in this case from 40 gigabytes（千兆字节）— worth of online writing. The version Mr. Seabrook tested was polished with back-issues of the New Yorker.

The metaphor of the brain is attractive, but “neurons” and “synapses” deserve those scare-quotes     ( 双引号). The system is merely making some—admittedly very complex—statistical guesses about which words follow which in a New Yorker-style sentence.

What escapes computers is creativity. They cannot create a topic or goal on their own, much less plan how to get there with logic and style. At various points in the online version of his article, readers can see how GPT-2 would have carried on writing Mr. Seabrook’s piece for him. The writing gives the impression of being human. But on closer inspection it is empty, even incoherent.

Meaningless writing is the preserve of artificial intelligence. But to truly write, you must first have something to say. Computers do not. They await instructions. To compose meaningful  essays, the likes of GPT-2 will first have to be integrated with databases of real-world knowledge.

1.Why does the author mention Gmail in the first paragraph?

A.To prepare for the introduction of the topic. B.To advertise the powerful Gmail.

C.To describe the feature of Gmail. D.To prove the strength of the Gmail.

2.In what way is GPT-2 more powerful than Gmail?

A.It is creative. B.It has neurons with synapses.

C.It can offer quick answers. D.It can follow a certain writing style.

3.What can we learn from the last two paragraphs?

A.GPT-2 is powerful because it is integrated to real-world knowledge.

B.GPT-2 has already been able to write meaningful essays.

C.GPT-2 still has a long way to write like humans.

D.GPT-2 can help Mr. Seabrook with writing very well.

4.Where is the passage taken from?

A.A brochure. B.An advertisement.

C.A magazine. D.An essay.

Answer

1.A

2.D

3.C

4.C

【解析】

本文是说明文，介绍了计算机自动写作功能的更强大的版本——GPT-2，它可以巧妙地模仿杂志的风格，但是写出来的文章， 如果仔细检查，就会觉得空洞，甚至是不连贯的。要写出有意义的文章，像GPT-2这类的项目首先必须与现实世界的知识数据库相结合。机器写手无法完全取代人类。

1.推理判断题。根据第二段“The New Yorker’s John Seabrook recently described a more powerful version of this technology, called GPT-2（纽约人John Seabrook最近描述了这项技术的一个更强大的版本，称为GPT-2）”可以推断，文章第1段中举出Gmail自动写作技术的例子，是为了引出下文关于GPT-2的话题。故选A项。

2.细节理解题。根据第二段“…. a more powerful version of this technology, called GPT-2,which can ably mimic (模拟) his magazine’s style”可知，计算机自动写作功能更强大的版本GPT-2可以巧妙地模仿杂志的风格。所以GPT-2可以遵循一定的写作风格方面比Gmail更强大。故选D项。

3.推理判断题。根据第五段“To compose meaningful essays, the likes of GPT-2 will first have to be integrated with databases of real-world knowledge（要写出有意义的文章，像GPT-2这样的项目首先必须与真实世界的知识数据库相结合）”可以推断，GPT-2技术要像人那样写出有意义的文章，还需要更多的研究，还有很长的路要走。故选C项。

4.推理判断题。本文是关于电脑自动写作功能更强大的版本——GPT-2的介绍，是关于人工智能研究的话题，应该是来自于科学杂志。故选C项。